Applied Intuition believes the next competitive edge in artificial intelligence won’t come from building bigger models, but from creating systems that help AI learn and improve faster in the real world.

In a recent post, co-founder and CTO Peter Ludwig argued that physical AI, from autonomous vehicles and warehouse robots to defense systems and industrial machines, depends less on raw model intelligence and more on engineering infrastructure that enables continuous testing, deployment, and improvement.

As AI models become more widely available, Ludwig said the advantage will shift to companies that can operationalize them at scale through rapid feedback loops, simulations, safety validation, and real-world data collection. “Intelligence is becoming ubiquitous. The ability to operationalize it isn’t,” he wrote.

Applied Intuition has been building that infrastructure through its AI engineering platform, Dana. The company says engineers have created more than 1,000 internal AI applications that have accelerated development workflows by roughly 20 times, allowing some software updates to be deployed multiple times a day instead of over weeks.

The company argues that while AI models alone don’t compound in value, systems do — because every deployment generates new data that improves future performance. As AI expands into transportation, robotics, manufacturing, and defense, Applied Intuition believes engineering platforms and continuous learning systems will become the industry’s next major moat.

Photo: Shutterstock